diff --git a/R/arguments.R b/R/arguments.R
index 6b14401bb..6ff64c6d5 100644
--- a/R/arguments.R
+++ b/R/arguments.R
@@ -86,7 +86,7 @@ check_others <- function(args, obj, core_args) {
 #'
 #' @export
 set_args <- function(object, ...) {
-  the_dots <- list(...)
+  the_dots <- enquos(...)
   if (length(the_dots) == 0)
     stop("Please pass at least one named argument.", call. = FALSE)
   main_args <- names(object$args)
diff --git a/R/boost_tree.R b/R/boost_tree.R
index 034390d6b..e6c309cc1 100644
--- a/R/boost_tree.R
+++ b/R/boost_tree.R
@@ -22,7 +22,7 @@
 #' }
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the  `...` slot. If left to their defaults
 #'  here (`NULL`), the values are taken from the underlying model
 #'  functions.  If parameters need to be modified, `update` can be used
 #'  in lieu of recreating the object from scratch.
@@ -30,8 +30,6 @@
 #' @param mode A single character string for the type of model.
 #'  Possible values for this model are "unknown", "regression", or
 #'  "classification".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `xgboost::xgb.train`, etc.). .
 #' @param mtry An number for the number (or proportion) of predictors that will
 #'  be randomly sampled at each split when creating the tree models (`xgboost`
 #'  only).
@@ -48,8 +46,11 @@
 #' @param sample_size An number for the number (or proportion) of data that is
 #'  exposed to the fitting routine. For `xgboost`, the sampling is done at at
 #'  each iteration while `C5.0` samples once during traning.
-#' @param ... Used for method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
+#' @param ... Other arguments to pass to the specific engine's
+#'  model fit function (see the Engine Details section below). This
+#'  should not include arguments defined by the main parameters to
+#'  this function. For the `update` function, the ellipses can
+#'  contain the primary arguments or any others.
 #' @details
 #' The data given to the function are not saved and are only used
 #'  to determine the _mode_ of the model. For `boost_tree`, the
@@ -62,12 +63,15 @@
 #' \item \pkg{Spark}: `"spark"`
 #' }
 #'
-#' Main parameter arguments (and those in `others`) can avoid
+#' Main parameter arguments (and those in `...`) can avoid
 #'  evaluation until the underlying function is executed by wrapping the
 #'  argument in [rlang::expr()] (e.g. `mtry = expr(floor(sqrt(p)))`).
 #'
+#'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
 #'  model, the template of the fit calls are:
 #'
@@ -114,13 +118,18 @@
 
 boost_tree <-
   function(mode = "unknown",
-           ...,
            mtry = NULL, trees = NULL, min_n = NULL,
            tree_depth = NULL, learn_rate = NULL,
            loss_reduction = NULL,
            sample_size = NULL,
-           others = list()) {
-    check_empty_ellipse(...)
+           ...) {
+    others         <- enquos(...)
+    mtry           <- enquo(mtry)
+    trees          <- enquo(trees)
+    min_n          <- enquo(min_n)
+    learn_rate     <- enquo(learn_rate)
+    loss_reduction <- enquo(loss_reduction)
+    sample_size    <- enquo(sample_size)
 
     if (!(mode %in% boost_tree_modes))
       stop("`mode` should be one of: ",
@@ -184,10 +193,15 @@ update.boost_tree <-
            mtry = NULL, trees = NULL, min_n = NULL,
            tree_depth = NULL, learn_rate = NULL,
            loss_reduction = NULL, sample_size = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others         <- enquos(...)
+    mtry           <- enquo(mtry)
+    trees          <- enquo(trees)
+    min_n          <- enquo(min_n)
+    learn_rate     <- enquo(learn_rate)
+    loss_reduction <- enquo(loss_reduction)
+    sample_size    <- enquo(sample_size)
 
     args <- list(
       mtry = mtry, trees = trees, min_n = min_n, tree_depth = tree_depth,
diff --git a/R/linear_reg.R b/R/linear_reg.R
index d2aed4342..c04ba78f3 100644
--- a/R/linear_reg.R
+++ b/R/linear_reg.R
@@ -12,25 +12,19 @@
 #' }
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the  `...` slot. If left to their defaults
 #'  here (`NULL`), the values are taken from the underlying model
 #'  functions. If parameters need to be modified, `update` can be used
 #'  in lieu of recreating the object from scratch.
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #'  The only possible value for this model is "regression".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `stats::lm`,
-#'  `rstanarm::stan_glm`, etc.). These are not evaluated
-#'  until the model is fit and will be substituted into the model
-#'  fit expression.
 #' @param penalty An non-negative number representing the
 #'  total amount of regularization (`glmnet` and `spark` only).
 #' @param mixture A number between zero and one (inclusive) that
 #'  represents the proportion of regularization that is used for the
 #'  L2 penalty (i.e. weight decay, or ridge regression) versus L1
 #'  (the lasso) (`glmnet` and `spark` only).
-#' @param ... Used for S3 method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
 #'
 #' @details
 #' The data given to the function are not saved and are only used
@@ -45,8 +39,10 @@
 #' \item \pkg{Spark}: `"spark"`
 #' }
 #'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
 #'  model, the template of the fit calls are:
 #'
@@ -105,11 +101,13 @@
 #' @importFrom purrr map_lgl
 linear_reg <-
   function(mode = "regression",
-           ...,
            penalty = NULL,
            mixture = NULL,
-           others = list()) {
-    check_empty_ellipse(...)
+           ...) {
+    others  <- enquos(...)
+    penalty <- enquo(penalty)
+    mixture <- enquo(mixture)
+
     if (!(mode %in% linear_reg_modes))
       stop(
         "`mode` should be one of: ",
@@ -121,7 +119,7 @@ linear_reg <-
       stop("The amount of regularization should be >= 0", call. = FALSE)
     if (is.numeric(mixture) && (mixture < 0 | mixture > 1))
       stop("The mixture proportion should be within [0,1]", call. = FALSE)
-    if (length(mixture) > 1)
+    if (is.numeric(mixture) && length(mixture) > 1)
       stop("Only one value of `mixture` is allowed.", call. = FALSE)
 
     args <- list(penalty = penalty, mixture = mixture)
@@ -156,11 +154,8 @@ print.linear_reg <- function(x, ...) {
 
 # ------------------------------------------------------------------------------
 
-#' @inheritParams linear_reg
+#' @inheritParams update.boost_tree
 #' @param object A linear regression model specification.
-#' @param fresh A logical for whether the arguments should be
-#'  modified in-place of or replaced wholesale.
-#' @return An updated model specification.
 #' @examples
 #' model <- linear_reg(penalty = 10, mixture = 0.1)
 #' model
@@ -172,10 +167,11 @@ print.linear_reg <- function(x, ...) {
 update.linear_reg <-
   function(object,
            penalty = NULL, mixture = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others  <- enquos(...)
+    penalty <- enquo(penalty)
+    mixture <- enquo(mixture)
 
     if (is.numeric(penalty) && penalty < 0)
       stop("The amount of regularization should be >= 0", call. = FALSE)
diff --git a/R/logistic_reg.R b/R/logistic_reg.R
index 7051b46a6..7533c916c 100644
--- a/R/logistic_reg.R
+++ b/R/logistic_reg.R
@@ -12,25 +12,19 @@
 #' }
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the  `...` slot. If left to their defaults
 #'  here (`NULL`), the values are taken from the underlying model
 #'  functions. If parameters need to be modified, `update` can be used
 #'  in lieu of recreating the object from scratch.
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #'  The only possible value for this model is "classification".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `stats::glm`,
-#'  `rstanarm::stan_glm`, etc.). These are not evaluated
-#'  until the model is fit and will be substituted into the model
-#'  fit expression.
 #' @param penalty An non-negative number representing the
 #'  total amount of regularization (`glmnet` and `spark` only).
 #' @param mixture A number between zero and one (inclusive) that
 #'  represents the proportion of regularization that is used for the
 #'  L2 penalty (i.e. weight decay, or ridge regression) versus L1
 #'  (the lasso) (`glmnet` and `spark` only).
-#' @param ... Used for S3 method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
 #' @details
 #' For `logistic_reg`, the mode will always be "classification".
 #'
@@ -42,8 +36,10 @@
 #' \item \pkg{Spark}: `"spark"`
 #' }
 #'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
 #'  model, the template of the fit calls are:
 #'
@@ -103,11 +99,13 @@
 #' @importFrom purrr map_lgl
 logistic_reg <-
   function(mode = "classification",
-           ...,
            penalty = NULL,
            mixture = NULL,
-           others = list()) {
-    check_empty_ellipse(...)
+           ...) {
+    others  <- enquos(...)
+    penalty <- enquo(penalty)
+    mixture <- enquo(mixture)
+
     if (!(mode %in% logistic_reg_modes))
       stop(
         "`mode` should be one of: ",
@@ -152,11 +150,8 @@ print.logistic_reg <- function(x, ...) {
 
 # ------------------------------------------------------------------------------
 
-#' @inheritParams logistic_reg
+#' @inheritParams update.boost_tree
 #' @param object A logistic regression model specification.
-#' @param fresh A logical for whether the arguments should be
-#'  modified in-place of or replaced wholesale.
-#' @return An updated model specification.
 #' @examples
 #' model <- logistic_reg(penalty = 10, mixture = 0.1)
 #' model
@@ -168,10 +163,11 @@ print.logistic_reg <- function(x, ...) {
 update.logistic_reg <-
   function(object,
            penalty = NULL, mixture = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others  <- enquos(...)
+    penalty <- enquo(penalty)
+    mixture <- enquo(mixture)
 
     if (is.numeric(penalty) && penalty < 0)
       stop("The amount of regularization should be >= 0", call. = FALSE)
diff --git a/R/logistic_reg_data.R b/R/logistic_reg_data.R
index 707d5a4c4..fb81e3353 100644
--- a/R/logistic_reg_data.R
+++ b/R/logistic_reg_data.R
@@ -30,7 +30,7 @@ logistic_reg_glm_data <-
       func = c(pkg = "stats", fun = "glm"),
       defaults =
         list(
-          family = expr(binomial)
+          family = expr(stats::binomial)
         )
     ),
     classes = list(
@@ -151,7 +151,7 @@ logistic_reg_stan_data <-
       func = c(pkg = "rstanarm", fun = "stan_glm"),
       defaults =
         list(
-          family = expr(binomial)
+          family = expr(stats::binomial)
         )
     ),
     classes = list(
diff --git a/R/mars.R b/R/mars.R
index dbaa8e381..cf153d139 100644
--- a/R/mars.R
+++ b/R/mars.R
@@ -17,26 +17,20 @@
 #' }
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the  `...` slot. If left to their defaults
 #'  here (`NULL`), the values are taken from the underlying model
 #'  functions. If parameters need to be modified, `update` can be used
 #'  in lieu of recreating the object from scratch.
 #'
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #'  Possible values for this model are "unknown", "regression", or
 #'  "classification".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `earth::earth`, etc.). If the outcome is a factor
-#'  and `mode = "classification"`, `others` can include the `glm` argument to
-#'  `earth::earth`. If this argument is not passed, it will be added prior to
-#'  the fitting occurs.
 #' @param num_terms The number of features that will be retained in the
 #'    final model, including the intercept.
 #' @param prod_degree The highest possible interaction degree.
 #' @param prune_method The pruning method.
-#' @param ... Used for method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
-#' @details Main parameter arguments (and those in `others`) can avoid
+#' @details Main parameter arguments (and those in `...`) can avoid
 #'  evaluation until the underlying function is executed by wrapping the
 #'  argument in [rlang::expr()].
 #'
@@ -46,8 +40,10 @@
 #' \item \pkg{R}:  `"earth"`
 #' }
 #'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
 #'  model, the template of the fit calls are:
 #'
@@ -71,10 +67,12 @@
 
 mars <-
   function(mode = "unknown",
-           ...,
            num_terms = NULL, prod_degree = NULL, prune_method = NULL,
-           others = list()) {
-    check_empty_ellipse(...)
+           ...) {
+    others  <- enquos(...)
+    num_terms    <- enquo(num_terms)
+    prod_degree  <- enquo(prod_degree)
+    prune_method <- enquo(prune_method)
 
     if (!(mode %in% mars_modes))
       stop("`mode` should be one of: ",
@@ -87,7 +85,7 @@ mars <-
       stop("`num_terms` should be >= 1", call. = FALSE)
     if (!is_varying(prune_method) &&
         !is.null(prune_method) &&
-        !is.character(prune_method))
+         is.character(prune_method))
       stop("`prune_method` should be a single string value", call. = FALSE)
 
     args <- list(num_terms = num_terms,
@@ -118,11 +116,8 @@ print.mars <- function(x, ...) {
 # ------------------------------------------------------------------------------
 
 #' @export
-#' @inheritParams mars
+#' @inheritParams update.boost_tree
 #' @param object A MARS model specification.
-#' @param fresh A logical for whether the arguments should be
-#'  modified in-place of or replaced wholesale.
-#' @return An updated model specification.
 #' @examples
 #' model <- mars(num_terms = 10, prune_method = "none")
 #' model
@@ -134,10 +129,12 @@ print.mars <- function(x, ...) {
 update.mars <-
   function(object,
            num_terms = NULL, prod_degree = NULL, prune_method = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others  <- enquos(...)
+    num_terms    <- enquo(num_terms)
+    prod_degree  <- enquo(prod_degree)
+    prune_method <- enquo(prune_method)
 
     args <- list(num_terms = num_terms,
                  prod_degree = prod_degree,
diff --git a/R/mlp.R b/R/mlp.R
index e4c3df660..8934318dd 100644
--- a/R/mlp.R
+++ b/R/mlp.R
@@ -18,7 +18,7 @@
 #'
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the  `...` slot. If left to their defaults
 #'  here (see above), the values are taken from the underlying model
 #'  functions. One exception is `hidden_units` when `nnet::nnet` is used; that
 #'  function's `size` argument has no default so a value of 5 units will be
@@ -26,13 +26,11 @@
 #'  `nnet::nnet` will be set to `TRUE` when a regression model is created.
 #'  If parameters need to be modified, `update` can be used
 #'  in lieu of recreating the object from scratch.
-
+#'
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #'  Possible values for this model are "unknown", "regression", or
 #'  "classification".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `nnet::nnet`,
-#'  `keras::fit`, `keras::compile`, etc.). .
 #' @param hidden_units An integer for the number of units in the hidden model.
 #' @param penalty A non-negative numeric value for the amount of weight
 #'  decay.
@@ -44,8 +42,6 @@
 #'  function between the hidden and output layers is automatically set to either
 #'  "linear" or "softmax" depending on the type of outcome. Possible values are:
 #'  "linear", "softmax", "relu", and "elu"
-#' @param ... Used for method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
 #' @details
 #'
 #' The model can be created using the `fit()` function using the
@@ -55,15 +51,17 @@
 #' \item \pkg{keras}: `"keras"`
 #' }
 #'
-#' Main parameter arguments (and those in `others`) can avoid
+#' Main parameter arguments (and those in `...`) can avoid
 #'  evaluation until the underlying function is executed by wrapping the
 #'  argument in [rlang::expr()] (e.g. `hidden_units = expr(num_preds * 2)`).
 #'
 #'  An error is thrown if both `penalty` and `dropout` are specified for
 #'  `keras` models.
 #'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
 #'  model, the template of the fit calls are:
 #'
@@ -93,11 +91,16 @@
 
 mlp <-
   function(mode = "unknown",
-           ...,
            hidden_units = NULL, penalty = NULL, dropout = NULL, epochs = NULL,
            activation = NULL,
-           others = list()) {
-    check_empty_ellipse(...)
+           ...) {
+    others  <- enquos(...)
+    hidden_units <- enquo(hidden_units)
+    penalty      <- enquo(penalty)
+    dropout      <- enquo(dropout)
+    epochs       <- enquo(epochs)
+    activation   <- enquo(activation)
+
 
     act_funs <- c("linear", "softmax", "relu", "elu")
     if (is.numeric(hidden_units))
@@ -157,11 +160,8 @@ print.mlp <- function(x, ...) {
 #'  in lieu of recreating the object from scratch.
 #'
 #' @export
-#' @inheritParams mlp
+#' @inheritParams update.boost_tree
 #' @param object A random forest model specification.
-#' @param fresh A logical for whether the arguments should be
-#'  modified in-place of or replaced wholesale.
-#' @return An updated model specification.
 #' @examples
 #' model <- mlp(hidden_units = 10, dropout = 0.30)
 #' model
@@ -174,10 +174,14 @@ update.mlp <-
   function(object,
            hidden_units = NULL, penalty = NULL, dropout = NULL,
            epochs = NULL, activation = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others  <- enquos(...)
+    hidden_units <- enquo(hidden_units)
+    penalty      <- enquo(penalty)
+    dropout      <- enquo(dropout)
+    epochs       <- enquo(epochs)
+    activation   <- enquo(activation)
 
     args <- list(hidden_units = hidden_units, penalty = penalty, dropout = dropout,
                  epochs = epochs, activation = activation)
@@ -209,8 +213,9 @@ update.mlp <-
 translate.mlp <- function(x, engine, ...) {
 
   if (engine == "nnet") {
-    if(is.null(x$args$hidden_units))
+    if(isTRUE(is.null(quo_get_expr(x$args$hidden_units)))) {
       x$args$hidden_units <- 5
+    }
   }
 
   x <- translate.default(x, engine, ...)
diff --git a/R/multinom_reg.R b/R/multinom_reg.R
index 6f079f167..a9542fa78 100644
--- a/R/multinom_reg.R
+++ b/R/multinom_reg.R
@@ -12,24 +12,19 @@
 #' }
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the `...` slot. If left to their defaults
 #'  here (`NULL`), the values are taken from the underlying model
 #'  functions. If parameters need to be modified, `update` can be used
 #'  in lieu of recreating the object from scratch.
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #'  The only possible value for this model is "classification".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `glmnet::glmnet` etc.). These are not evaluated
-#'  until the model is fit and will be substituted into the model
-#'  fit expression.
 #' @param penalty An non-negative number representing the
 #'  total amount of regularization.
 #' @param mixture A number between zero and one (inclusive) that
 #'  represents the proportion of regularization that is used for the
 #'  L2 penalty (i.e. weight decay, or ridge regression) versus L1
 #'  (the lasso) (`glmnet` only).
-#' @param ... Used for S3 method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
 #' @details
 #' For `multinom_reg`, the mode will always be "classification".
 #'
@@ -40,8 +35,10 @@
 #' \item \pkg{Stan}:  `"stan"`
 #' }
 #'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
 #'  model, the template of the fit calls are:
 #'
@@ -85,11 +82,13 @@
 #' @importFrom purrr map_lgl
 multinom_reg <-
   function(mode = "classification",
-           ...,
            penalty = NULL,
            mixture = NULL,
-           others = list()) {
-    check_empty_ellipse(...)
+           ...) {
+    others  <- enquos(...)
+    penalty <- enquo(penalty)
+    mixture <- enquo(mixture)
+
     if (!(mode %in% multinom_reg_modes))
       stop(
         "`mode` should be one of: ",
@@ -134,11 +133,8 @@ print.multinom_reg <- function(x, ...) {
 
 # ------------------------------------------------------------------------------
 
-#' @inheritParams multinom_reg
+#' @inheritParams update.boost_tree
 #' @param object A multinomial regression model specification.
-#' @param fresh A logical for whether the arguments should be
-#'  modified in-place of or replaced wholesale.
-#' @return An updated model specification.
 #' @examples
 #' model <- multinom_reg(penalty = 10, mixture = 0.1)
 #' model
@@ -150,10 +146,11 @@ print.multinom_reg <- function(x, ...) {
 update.multinom_reg <-
   function(object,
            penalty = NULL, mixture = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others  <- enquos(...)
+    penalty <- enquo(penalty)
+    mixture <- enquo(mixture)
 
     if (is.numeric(penalty) && penalty < 0)
       stop("The amount of regularization should be >= 0", call. = FALSE)
diff --git a/R/nearest_neighbor.R b/R/nearest_neighbor.R
index 499be4ea0..ef638cb3d 100644
--- a/R/nearest_neighbor.R
+++ b/R/nearest_neighbor.R
@@ -19,11 +19,11 @@
 #' }
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the  `...` slot. If left to their defaults
 #'  here (`NULL`), the values are taken from the underlying model
 #'  functions. If parameters need to be modified, `update()` can be used
 #'  in lieu of recreating the object from scratch.
-#'
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #' Possible values for this model are `"unknown"`, `"regression"`, or
 #' `"classification"`.
@@ -39,14 +39,6 @@
 #' @param dist_power A single number for the parameter used in
 #' calculating Minkowski distance.
 #'
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `kknn::train.kknn`). These are not evaluated
-#'  until the model is fit and will be substituted into the model
-#'  fit expression.
-#'
-#' @param ... Used for S3 method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
-#'
 #' @details
 #' The model can be created using the `fit()` function using the
 #'  following _engines_:
@@ -54,8 +46,10 @@
 #' \item \pkg{R}:  `"kknn"`
 #' }
 #'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
 #'  model, the template of the fit calls are:
 #'
@@ -77,13 +71,14 @@
 #'
 #' @export
 nearest_neighbor <- function(mode = "unknown",
-                             ...,
                              neighbors = NULL,
                              weight_func = NULL,
                              dist_power = NULL,
-                             others = list()) {
-
-  check_empty_ellipse(...)
+                             ...) {
+  others      <- enquos(...)
+  neighbors   <- enquo(neighbors)
+  weight_func <- enquo(weight_func)
+  dist_power  <- enquo(dist_power)
 
   ## TODO: make a utility function here
   if (!(mode %in% nearest_neighbor_modes)) {
@@ -132,15 +127,18 @@ print.nearest_neighbor <- function(x, ...) {
 # ------------------------------------------------------------------------------
 
 #' @export
+#' @inheritParams update.boost_tree
 update.nearest_neighbor <- function(object,
                                     neighbors = NULL,
                                     weight_func = NULL,
                                     dist_power = NULL,
-                                    others = list(),
                                     fresh = FALSE,
                                     ...) {
 
-  check_empty_ellipse(...)
+  others      <- enquos(...)
+  neighbors   <- enquo(neighbors)
+  weight_func <- enquo(weight_func)
+  dist_power  <- enquo(dist_power)
 
   if(is.numeric(neighbors) && !positive_int_scalar(neighbors)) {
     stop("`neighbors` must be a length 1 positive integer.", call. = FALSE)
diff --git a/R/rand_forest.R b/R/rand_forest.R
index 53b01401b..6c89f4eb0 100644
--- a/R/rand_forest.R
+++ b/R/rand_forest.R
@@ -15,25 +15,21 @@
 #' }
 #' These arguments are converted to their specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to their defaults
+#'  set using the  `...` slot. If left to their defaults
 #'  here (`NULL`), the values are taken from the underlying model
 #'  functions. If parameters need to be modified, `update` can be used
 #'  in lieu of recreating the object from scratch.
 #'
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #'  Possible values for this model are "unknown", "regression", or
 #'  "classification".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `ranger::ranger`,
-#'  `randomForest::randomForest`, etc.). .
 #' @param mtry An integer for the number of predictors that will
 #'  be randomly sampled at each split when creating the tree models.
 #' @param trees An integer for the number of trees contained in
 #'  the ensemble.
 #' @param min_n An integer for the minimum number of data points
 #'  in a node that are required for the node to be split further.
-#' @param ... Used for method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
 #' @details
 #' The model can be created using the `fit()` function using the
 #'  following _engines_:
@@ -42,14 +38,16 @@
 #' \item \pkg{Spark}: `"spark"`
 #' }
 #'
-#' Main parameter arguments (and those in `others`) can avoid
+#' Main parameter arguments (and those in `...`) can avoid
 #'  evaluation until the underlying function is executed by wrapping the
 #'  argument in [rlang::expr()] (e.g. `mtry = expr(floor(sqrt(p)))`).
 #'
+#' @section Engine Details:
+#'
 #' Engines may have pre-set default arguments when executing the
-#'  model fit call. These can be changed by using the `others`
+#'  model fit call. These can be changed by using the `...`
 #'  argument to pass in the preferred values. For this type of
-#'  model, the template of the fit calls are:
+#'  model, the template of the fit calls are::
 #'
 #' \pkg{ranger} classification
 #'
@@ -144,11 +142,8 @@ print.rand_forest <- function(x, ...) {
 # ------------------------------------------------------------------------------
 
 #' @export
-#' @inheritParams rand_forest
+#' @inheritParams update.boost_tree
 #' @param object A random forest model specification.
-#' @param fresh A logical for whether the arguments should be
-#'  modified in-place of or replaced wholesale.
-#' @return An updated model specification.
 #' @examples
 #' model <- rand_forest(mtry = 10, min_n = 3)
 #' model
@@ -160,10 +155,12 @@ print.rand_forest <- function(x, ...) {
 update.rand_forest <-
   function(object,
            mtry = NULL, trees = NULL, min_n = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others <- enquos(...)
+    mtry   <- enquo(mtry)
+    trees  <- enquo(trees)
+    min_n  <- enquo(min_n)
 
     args <- list(mtry = mtry, trees = trees, min_n = min_n)
 
@@ -194,34 +191,42 @@ update.rand_forest <-
 translate.rand_forest <- function(x, engine, ...) {
   x <- translate.default(x, engine, ...)
 
+  # slightly cleaner code using
+  arg_vals <- x$method$fit$args
+
   if (x$engine == "spark") {
-    if (x$mode == "unknown")
+    if (x$mode == "unknown") {
       stop(
         "For spark random forests models, the mode cannot be 'unknown' ",
         "if the specification is to be translated.",
         call. = FALSE
       )
-    else
-      x$method$fit$args$type <- x$mode
-
-    # See "Details" in ?ml_random_forest_classifier
-    if (is.numeric(x$method$fit$args$feature_subset_strategy))
-      x$method$fit$args$feature_subset_strategy <-
-        paste(x$method$fit$args$feature_subset_strategy)
+    } else {
+      arg_vals$type <- x$mode
+    }
 
+    # See "Details" in ?ml_random_forest_classifier. `feature_subset_strategy`
+    # should be character even if it contains a number.
+    if (any(names(arg_vals) == "feature_subset_strategy") &&
+        isTRUE(is.numeric(quo_get_expr(arg_vals$feature_subset_strategy)))) {
+      arg_vals$feature_subset_strategy <-
+        paste(quo_get_expr(arg_vals$feature_subset_strategy))
+    }
   }
 
   # add checks to error trap or change things for this method
   if (x$engine == "ranger") {
-    if (any(names(x$method$fit$args) == "importance"))
-      if (is.logical(x$method$fit$args$importance))
+    if (any(names(arg_vals) == "importance"))
+      if (isTRUE(is.logical(quo_get_expr(arg_vals$importance))))
         stop("`importance` should be a character value. See ?ranger::ranger.",
              call. = FALSE)
     # unless otherwise specified, classification models are probability forests
-    if (x$mode == "classification" && !any(names(x$method$fit$args) == "probability"))
-      x$method$fit$args$probability <- TRUE
+    if (x$mode == "classification" && !any(names(arg_vals) == "probability"))
+      arg_vals$probability <- TRUE
 
   }
+  x$method$fit$args <- arg_vals
+
   x
 }
 
diff --git a/R/surv_reg.R b/R/surv_reg.R
index 16ad84b70..5e391c191 100644
--- a/R/surv_reg.R
+++ b/R/surv_reg.R
@@ -9,7 +9,7 @@
 #' }
 #' This argument is converted to its specific names at the
 #'  time that the model is fit. Other options and argument can be
-#'  set using the `others` argument. If left to its default
+#'  set using the  `...` slot. If left to its default
 #'  here (`NULL`), the value is taken from the underlying model
 #'  functions.
 #'
@@ -30,16 +30,11 @@
 #' \itemize{
 #' \item \pkg{R}:  `"flexsurv"`
 #' }
+#' @inheritParams boost_tree
 #' @param mode A single character string for the type of model.
 #'  The only possible value for this model is "regression".
-#' @param others A named list of arguments to be used by the
-#'  underlying models (e.g., `flexsurv::flexsurvreg`). These are not evaluated
-#'  until the model is fit and will be substituted into the model
-#'  fit expression.
 #' @param dist A character string for the outcome distribution. "weibull" is
 #'  the default.
-#' @param ... Used for S3 method consistency. Any arguments passed to
-#'  the ellipses will result in an error. Use `others` instead.
 #' @seealso [varying()], [fit()], [survival::Surv()]
 #' @references Jackson, C. (2016). `flexsurv`: A Platform for Parametric Survival
 #'  Modeling in R. _Journal of Statistical Software_, 70(8), 1 - 33.
@@ -51,10 +46,11 @@
 #' @export
 surv_reg <-
   function(mode = "regression",
-           ...,
            dist = NULL,
-           others = list()) {
-    check_empty_ellipse(...)
+           ...) {
+    others <- enquos(...)
+    dist   <- enquo(dist)
+
     if (!(mode %in% surv_reg_modes))
       stop(
         "`mode` should be one of: ",
@@ -98,11 +94,8 @@ print.surv_reg <- function(x, ...) {
 #' If parameters need to be modified, this function can be used
 #'  in lieu of recreating the object from scratch.
 #'
-#' @inheritParams surv_reg
+#' @inheritParams update.boost_tree
 #' @param object A survival regression model specification.
-#' @param fresh A logical for whether the arguments should be
-#'  modified in-place of or replaced wholesale.
-#' @return An updated model specification.
 #' @examples
 #' model <- surv_reg(dist = "weibull")
 #' model
@@ -113,10 +106,10 @@ print.surv_reg <- function(x, ...) {
 update.surv_reg <-
   function(object,
            dist = NULL,
-           others = list(),
            fresh = FALSE,
            ...) {
-    check_empty_ellipse(...)
+    others <- enquos(...)
+    dist   <- enquo(dist)
 
     args <- list(dist = dist)
 
diff --git a/man/boost_tree.Rd b/man/boost_tree.Rd
index a7520d251..1e55def1d 100644
--- a/man/boost_tree.Rd
+++ b/man/boost_tree.Rd
@@ -5,23 +5,19 @@
 \alias{update.boost_tree}
 \title{General Interface for Boosted Trees}
 \usage{
-boost_tree(mode = "unknown", ..., mtry = NULL, trees = NULL,
+boost_tree(mode = "unknown", mtry = NULL, trees = NULL,
   min_n = NULL, tree_depth = NULL, learn_rate = NULL,
-  loss_reduction = NULL, sample_size = NULL, others = list())
+  loss_reduction = NULL, sample_size = NULL, ...)
 
 \method{update}{boost_tree}(object, mtry = NULL, trees = NULL,
   min_n = NULL, tree_depth = NULL, learn_rate = NULL,
-  loss_reduction = NULL, sample_size = NULL, others = list(),
-  fresh = FALSE, ...)
+  loss_reduction = NULL, sample_size = NULL, fresh = FALSE, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 Possible values for this model are "unknown", "regression", or
 "classification".}
 
-\item{...}{Used for method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{mtry}{An number for the number (or proportion) of predictors that will
 be randomly sampled at each split when creating the tree models (\code{xgboost}
 only).}
@@ -45,8 +41,11 @@ to split further  (\code{xgboost} only).}
 exposed to the fitting routine. For \code{xgboost}, the sampling is done at at
 each iteration while \code{C5.0} samples once during traning.}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{xgboost::xgb.train}, etc.). .}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A boosted tree model specification.}
 
@@ -77,7 +76,7 @@ to split further.
 }
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the  \code{...} slot. If left to their defaults
 here (\code{NULL}), the values are taken from the underlying model
 functions.  If parameters need to be modified, \code{update} can be used
 in lieu of recreating the object from scratch.
@@ -94,12 +93,29 @@ following \emph{engines}:
 \item \pkg{Spark}: \code{"spark"}
 }
 
-Main parameter arguments (and those in \code{others}) can avoid
+Main parameter arguments (and those in \code{...}) can avoid
 evaluation until the underlying function is executed by wrapping the
 argument in \code{\link[rlang:expr]{rlang::expr()}} (e.g. \code{mtry = expr(floor(sqrt(p)))}).
+}
+\note{
+For models created using the spark engine, there are
+several differences to consider. First, only the formula
+interface to via \code{fit} is available; using \code{fit_xy} will
+generate an error. Second, the predictions will always be in a
+spark table format. The names will be the same as documented but
+without the dots. Third, there is no equivalent to factor
+columns in spark tables so class predictions are returned as
+character columns. Fourth, to retain the model object for a new
+R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
+object should be serialized via \code{ml_save(object$fit)} and
+separately saved to disk. In a new session, the object can be
+reloaded and reattached to the \code{parsnip} object.
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
 model, the template of the fit calls are:
 
@@ -123,20 +139,7 @@ model, the template of the fit calls are:
 
 \Sexpr[results=rd]{parsnip:::show_fit(parsnip:::boost_tree(mode = "regression"), "spark")}
 }
-\note{
-For models created using the spark engine, there are
-several differences to consider. First, only the formula
-interface to via \code{fit} is available; using \code{fit_xy} will
-generate an error. Second, the predictions will always be in a
-spark table format. The names will be the same as documented but
-without the dots. Third, there is no equivalent to factor
-columns in spark tables so class predictions are returned as
-character columns. Fourth, to retain the model object for a new
-R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
-object should be serialized via \code{ml_save(object$fit)} and
-separately saved to disk. In a new session, the object can be
-reloaded and reattached to the \code{parsnip} object.
-}
+
 \examples{
 boost_tree(mode = "classification", trees = 20)
 # Parameters can be represented by a placeholder:
diff --git a/man/linear_reg.Rd b/man/linear_reg.Rd
index b108de728..e227b9796 100644
--- a/man/linear_reg.Rd
+++ b/man/linear_reg.Rd
@@ -5,19 +5,15 @@
 \alias{update.linear_reg}
 \title{General Interface for Linear Regression Models}
 \usage{
-linear_reg(mode = "regression", ..., penalty = NULL, mixture = NULL,
-  others = list())
+linear_reg(mode = "regression", penalty = NULL, mixture = NULL, ...)
 
 \method{update}{linear_reg}(object, penalty = NULL, mixture = NULL,
-  others = list(), fresh = FALSE, ...)
+  fresh = FALSE, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 The only possible value for this model is "regression".}
 
-\item{...}{Used for S3 method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{penalty}{An non-negative number representing the
 total amount of regularization (\code{glmnet} and \code{spark} only).}
 
@@ -26,20 +22,17 @@ represents the proportion of regularization that is used for the
 L2 penalty (i.e. weight decay, or ridge regression) versus L1
 (the lasso) (\code{glmnet} and \code{spark} only).}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{stats::lm},
-\code{rstanarm::stan_glm}, etc.). These are not evaluated
-until the model is fit and will be substituted into the model
-fit expression.}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A linear regression model specification.}
 
 \item{fresh}{A logical for whether the arguments should be
 modified in-place of or replaced wholesale.}
 }
-\value{
-An updated model specification.
-}
 \description{
 \code{linear_reg} is a way to generate a \emph{specification} of a model
 before fitting and allows the model to be created using
@@ -53,7 +46,7 @@ the model. Note that this will be ignored for some engines.
 }
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the  \code{...} slot. If left to their defaults
 here (\code{NULL}), the values are taken from the underlying model
 functions. If parameters need to be modified, \code{update} can be used
 in lieu of recreating the object from scratch.
@@ -70,9 +63,26 @@ following \emph{engines}:
 \item \pkg{Stan}:  \code{"stan"}
 \item \pkg{Spark}: \code{"spark"}
 }
+}
+\note{
+For models created using the spark engine, there are
+several differences to consider. First, only the formula
+interface to via \code{fit} is available; using \code{fit_xy} will
+generate an error. Second, the predictions will always be in a
+spark table format. The names will be the same as documented but
+without the dots. Third, there is no equivalent to factor
+columns in spark tables so class predictions are returned as
+character columns. Fourth, to retain the model object for a new
+R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
+object should be serialized via \code{ml_save(object$fit)} and
+separately saved to disk. In a new session, the object can be
+reloaded and reattached to the \code{parsnip} object.
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
 model, the template of the fit calls are:
 
@@ -109,20 +119,7 @@ these instances, the units are the original outcome and when
 distribution (or posterior predictive distribution as
 appropriate) is returned.
 }
-\note{
-For models created using the spark engine, there are
-several differences to consider. First, only the formula
-interface to via \code{fit} is available; using \code{fit_xy} will
-generate an error. Second, the predictions will always be in a
-spark table format. The names will be the same as documented but
-without the dots. Third, there is no equivalent to factor
-columns in spark tables so class predictions are returned as
-character columns. Fourth, to retain the model object for a new
-R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
-object should be serialized via \code{ml_save(object$fit)} and
-separately saved to disk. In a new session, the object can be
-reloaded and reattached to the \code{parsnip} object.
-}
+
 \examples{
 linear_reg()
 # Parameters can be represented by a placeholder:
diff --git a/man/logistic_reg.Rd b/man/logistic_reg.Rd
index 1d4fc0533..d466ef684 100644
--- a/man/logistic_reg.Rd
+++ b/man/logistic_reg.Rd
@@ -5,19 +5,16 @@
 \alias{update.logistic_reg}
 \title{General Interface for Logistic Regression Models}
 \usage{
-logistic_reg(mode = "classification", ..., penalty = NULL,
-  mixture = NULL, others = list())
+logistic_reg(mode = "classification", penalty = NULL, mixture = NULL,
+  ...)
 
 \method{update}{logistic_reg}(object, penalty = NULL, mixture = NULL,
-  others = list(), fresh = FALSE, ...)
+  fresh = FALSE, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 The only possible value for this model is "classification".}
 
-\item{...}{Used for S3 method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{penalty}{An non-negative number representing the
 total amount of regularization (\code{glmnet} and \code{spark} only).}
 
@@ -26,20 +23,17 @@ represents the proportion of regularization that is used for the
 L2 penalty (i.e. weight decay, or ridge regression) versus L1
 (the lasso) (\code{glmnet} and \code{spark} only).}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{stats::glm},
-\code{rstanarm::stan_glm}, etc.). These are not evaluated
-until the model is fit and will be substituted into the model
-fit expression.}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A logistic regression model specification.}
 
 \item{fresh}{A logical for whether the arguments should be
 modified in-place of or replaced wholesale.}
 }
-\value{
-An updated model specification.
-}
 \description{
 \code{logistic_reg} is a way to generate a \emph{specification} of a model
 before fitting and allows the model to be created using
@@ -53,7 +47,7 @@ the model. Note that this will be ignored for some engines.
 }
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the  \code{...} slot. If left to their defaults
 here (\code{NULL}), the values are taken from the underlying model
 functions. If parameters need to be modified, \code{update} can be used
 in lieu of recreating the object from scratch.
@@ -68,9 +62,26 @@ following \emph{engines}:
 \item \pkg{Stan}:  \code{"stan"}
 \item \pkg{Spark}: \code{"spark"}
 }
+}
+\note{
+For models created using the spark engine, there are
+several differences to consider. First, only the formula
+interface to via \code{fit} is available; using \code{fit_xy} will
+generate an error. Second, the predictions will always be in a
+spark table format. The names will be the same as documented but
+without the dots. Third, there is no equivalent to factor
+columns in spark tables so class predictions are returned as
+character columns. Fourth, to retain the model object for a new
+R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
+object should be serialized via \code{ml_save(object$fit)} and
+separately saved to disk. In a new session, the object can be
+reloaded and reattached to the \code{parsnip} object.
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
 model, the template of the fit calls are:
 
@@ -108,20 +119,7 @@ distribution (or posterior predictive distribution as
 appropriate) is returned. For \code{glm}, the standard error is in logit units
 while the intervals are in probability units.
 }
-\note{
-For models created using the spark engine, there are
-several differences to consider. First, only the formula
-interface to via \code{fit} is available; using \code{fit_xy} will
-generate an error. Second, the predictions will always be in a
-spark table format. The names will be the same as documented but
-without the dots. Third, there is no equivalent to factor
-columns in spark tables so class predictions are returned as
-character columns. Fourth, to retain the model object for a new
-R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
-object should be serialized via \code{ml_save(object$fit)} and
-separately saved to disk. In a new session, the object can be
-reloaded and reattached to the \code{parsnip} object.
-}
+
 \examples{
 logistic_reg()
 # Parameters can be represented by a placeholder:
diff --git a/man/mars.Rd b/man/mars.Rd
index f19dbc139..9f4d25e03 100644
--- a/man/mars.Rd
+++ b/man/mars.Rd
@@ -5,20 +5,17 @@
 \alias{update.mars}
 \title{General Interface for MARS}
 \usage{
-mars(mode = "unknown", ..., num_terms = NULL, prod_degree = NULL,
-  prune_method = NULL, others = list())
+mars(mode = "unknown", num_terms = NULL, prod_degree = NULL,
+  prune_method = NULL, ...)
 
 \method{update}{mars}(object, num_terms = NULL, prod_degree = NULL,
-  prune_method = NULL, others = list(), fresh = FALSE, ...)
+  prune_method = NULL, fresh = FALSE, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 Possible values for this model are "unknown", "regression", or
 "classification".}
 
-\item{...}{Used for method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{num_terms}{The number of features that will be retained in the
 final model, including the intercept.}
 
@@ -26,20 +23,17 @@ final model, including the intercept.}
 
 \item{prune_method}{The pruning method.}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{earth::earth}, etc.). If the outcome is a factor
-and \code{mode = "classification"}, \code{others} can include the \code{glm} argument to
-\code{earth::earth}. If this argument is not passed, it will be added prior to
-the fitting occurs.}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A MARS model specification.}
 
 \item{fresh}{A logical for whether the arguments should be
 modified in-place of or replaced wholesale.}
 }
-\value{
-An updated model specification.
-}
 \description{
 \code{mars} is a way to generate a \emph{specification} of a model before
 fitting and allows the model to be created using R. The main
@@ -56,13 +50,13 @@ in \code{?earth}.
 }
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the  \code{...} slot. If left to their defaults
 here (\code{NULL}), the values are taken from the underlying model
 functions. If parameters need to be modified, \code{update} can be used
 in lieu of recreating the object from scratch.
 }
 \details{
-Main parameter arguments (and those in \code{others}) can avoid
+Main parameter arguments (and those in \code{...}) can avoid
 evaluation until the underlying function is executed by wrapping the
 argument in \code{\link[rlang:expr]{rlang::expr()}}.
 
@@ -71,9 +65,12 @@ following \emph{engines}:
 \itemize{
 \item \pkg{R}:  \code{"earth"}
 }
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
 model, the template of the fit calls are:
 
@@ -89,6 +86,7 @@ Note that, when the model is fit, the \pkg{earth} package only has its
 namespace loaded. However, if \code{multi_predict} is used, the package is
 attached.
 }
+
 \examples{
 mars(mode = "regression", num_terms = 5)
 model <- mars(num_terms = 10, prune_method = "none")
diff --git a/man/mlp.Rd b/man/mlp.Rd
index 437e93f79..807fd04ae 100644
--- a/man/mlp.Rd
+++ b/man/mlp.Rd
@@ -5,22 +5,18 @@
 \alias{update.mlp}
 \title{General Interface for Single Layer Neural Network}
 \usage{
-mlp(mode = "unknown", ..., hidden_units = NULL, penalty = NULL,
-  dropout = NULL, epochs = NULL, activation = NULL,
-  others = list())
+mlp(mode = "unknown", hidden_units = NULL, penalty = NULL,
+  dropout = NULL, epochs = NULL, activation = NULL, ...)
 
 \method{update}{mlp}(object, hidden_units = NULL, penalty = NULL,
-  dropout = NULL, epochs = NULL, activation = NULL,
-  others = list(), fresh = FALSE, ...)
+  dropout = NULL, epochs = NULL, activation = NULL, fresh = FALSE,
+  ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 Possible values for this model are "unknown", "regression", or
 "classification".}
 
-\item{...}{Used for method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{hidden_units}{An integer for the number of units in the hidden model.}
 
 \item{penalty}{A non-negative numeric value for the amount of weight
@@ -37,18 +33,17 @@ function between the hidden and output layers is automatically set to either
 "linear" or "softmax" depending on the type of outcome. Possible values are:
 "linear", "softmax", "relu", and "elu"}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{nnet::nnet},
-\code{keras::fit}, \code{keras::compile}, etc.). .}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A random forest model specification.}
 
 \item{fresh}{A logical for whether the arguments should be
 modified in-place of or replaced wholesale.}
 }
-\value{
-An updated model specification.
-}
 \description{
 \code{mlp}, for multilayer perceptron, is a way to generate a \emph{specification} of
 a model before fitting and allows the model to be created using
@@ -72,7 +67,7 @@ in lieu of recreating the object from scratch.
 \details{
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the  \code{...} slot. If left to their defaults
 here (see above), the values are taken from the underlying model
 functions. One exception is \code{hidden_units} when \code{nnet::nnet} is used; that
 function's \code{size} argument has no default so a value of 5 units will be
@@ -88,15 +83,18 @@ following \emph{engines}:
 \item \pkg{keras}: \code{"keras"}
 }
 
-Main parameter arguments (and those in \code{others}) can avoid
+Main parameter arguments (and those in \code{...}) can avoid
 evaluation until the underlying function is executed by wrapping the
 argument in \code{\link[rlang:expr]{rlang::expr()}} (e.g. \code{hidden_units = expr(num_preds * 2)}).
 
 An error is thrown if both \code{penalty} and \code{dropout} are specified for
 \code{keras} models.
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
 model, the template of the fit calls are:
 
@@ -116,6 +114,7 @@ model, the template of the fit calls are:
 
 \Sexpr[results=rd]{parsnip:::show_fit(parsnip:::mlp(mode = "regression"), "nnet")}
 }
+
 \examples{
 mlp(mode = "classification", penalty = 0.01)
 # Parameters can be represented by a placeholder:
diff --git a/man/multinom_reg.Rd b/man/multinom_reg.Rd
index 91a650952..db9ba3614 100644
--- a/man/multinom_reg.Rd
+++ b/man/multinom_reg.Rd
@@ -5,19 +5,16 @@
 \alias{update.multinom_reg}
 \title{General Interface for Multinomial Regression Models}
 \usage{
-multinom_reg(mode = "classification", ..., penalty = NULL,
-  mixture = NULL, others = list())
+multinom_reg(mode = "classification", penalty = NULL, mixture = NULL,
+  ...)
 
 \method{update}{multinom_reg}(object, penalty = NULL, mixture = NULL,
-  others = list(), fresh = FALSE, ...)
+  fresh = FALSE, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 The only possible value for this model is "classification".}
 
-\item{...}{Used for S3 method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{penalty}{An non-negative number representing the
 total amount of regularization.}
 
@@ -26,19 +23,17 @@ represents the proportion of regularization that is used for the
 L2 penalty (i.e. weight decay, or ridge regression) versus L1
 (the lasso) (\code{glmnet} only).}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{glmnet::glmnet} etc.). These are not evaluated
-until the model is fit and will be substituted into the model
-fit expression.}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A multinomial regression model specification.}
 
 \item{fresh}{A logical for whether the arguments should be
 modified in-place of or replaced wholesale.}
 }
-\value{
-An updated model specification.
-}
 \description{
 \code{multinom_reg} is a way to generate a \emph{specification} of a model
 before fitting and allows the model to be created using
@@ -52,7 +47,7 @@ the model. Note that this will be ignored for some engines.
 }
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the \code{...} slot. If left to their defaults
 here (\code{NULL}), the values are taken from the underlying model
 functions. If parameters need to be modified, \code{update} can be used
 in lieu of recreating the object from scratch.
@@ -66,9 +61,26 @@ following \emph{engines}:
 \item \pkg{R}:   \code{"glmnet"}
 \item \pkg{Stan}:  \code{"stan"}
 }
+}
+\note{
+For models created using the spark engine, there are
+several differences to consider. First, only the formula
+interface to via \code{fit} is available; using \code{fit_xy} will
+generate an error. Second, the predictions will always be in a
+spark table format. The names will be the same as documented but
+without the dots. Third, there is no equivalent to factor
+columns in spark tables so class predictions are returned as
+character columns. Fourth, to retain the model object for a new
+R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
+object should be serialized via \code{ml_save(object$fit)} and
+separately saved to disk. In a new session, the object can be
+reloaded and reattached to the \code{parsnip} object.
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
 model, the template of the fit calls are:
 
@@ -90,20 +102,7 @@ multiple values or no values for \code{penalty} are used in
 \code{multinom_reg}, the \code{predict} method will return a data frame with
 columns \code{values} and \code{lambda}.
 }
-\note{
-For models created using the spark engine, there are
-several differences to consider. First, only the formula
-interface to via \code{fit} is available; using \code{fit_xy} will
-generate an error. Second, the predictions will always be in a
-spark table format. The names will be the same as documented but
-without the dots. Third, there is no equivalent to factor
-columns in spark tables so class predictions are returned as
-character columns. Fourth, to retain the model object for a new
-R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
-object should be serialized via \code{ml_save(object$fit)} and
-separately saved to disk. In a new session, the object can be
-reloaded and reattached to the \code{parsnip} object.
-}
+
 \examples{
 multinom_reg()
 # Parameters can be represented by a placeholder:
diff --git a/man/nearest_neighbor.Rd b/man/nearest_neighbor.Rd
index 33bf3d34c..5851088c9 100644
--- a/man/nearest_neighbor.Rd
+++ b/man/nearest_neighbor.Rd
@@ -4,17 +4,14 @@
 \alias{nearest_neighbor}
 \title{General Interface for K-Nearest Neighbor Models}
 \usage{
-nearest_neighbor(mode = "unknown", ..., neighbors = NULL,
-  weight_func = NULL, dist_power = NULL, others = list())
+nearest_neighbor(mode = "unknown", neighbors = NULL,
+  weight_func = NULL, dist_power = NULL, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 Possible values for this model are \code{"unknown"}, \code{"regression"}, or
 \code{"classification"}.}
 
-\item{...}{Used for S3 method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{neighbors}{A single integer for the number of neighbors
 to consider (often called \code{k}).}
 
@@ -26,10 +23,11 @@ to weight distances between samples. Valid choices are: \code{"rectangular"},
 \item{dist_power}{A single number for the parameter used in
 calculating Minkowski distance.}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{kknn::train.kknn}). These are not evaluated
-until the model is fit and will be substituted into the model
-fit expression.}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 }
 \description{
 \code{nearest_neighbor()} is a way to generate a \emph{specification} of a model
@@ -47,7 +45,7 @@ and the Euclidean distance with \code{dist_power = 2}.
 }
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the  \code{...} slot. If left to their defaults
 here (\code{NULL}), the values are taken from the underlying model
 functions. If parameters need to be modified, \code{update()} can be used
 in lieu of recreating the object from scratch.
@@ -58,9 +56,19 @@ following \emph{engines}:
 \itemize{
 \item \pkg{R}:  \code{"kknn"}
 }
+}
+\note{
+For \code{kknn}, the underlying modeling function used is a restricted
+version of \code{train.kknn()} and not \code{kknn()}. It is set up in this way so that
+\code{parsnip} can utilize the underlying \code{predict.train.kknn} method to predict
+on new data. This also means that a single value of that function's
+\code{kernel} argument (a.k.a \code{weight_func} here) can be supplied
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
 model, the template of the fit calls are:
 
@@ -68,13 +76,7 @@ model, the template of the fit calls are:
 
 \Sexpr[results=rd]{parsnip:::show_fit(parsnip:::nearest_neighbor(), "kknn")}
 }
-\note{
-For \code{kknn}, the underlying modeling function used is a restricted
-version of \code{train.kknn()} and not \code{kknn()}. It is set up in this way so that
-\code{parsnip} can utilize the underlying \code{predict.train.kknn} method to predict
-on new data. This also means that a single value of that function's
-\code{kernel} argument (a.k.a \code{weight_func} here) can be supplied
-}
+
 \examples{
 nearest_neighbor()
 
diff --git a/man/rand_forest.Rd b/man/rand_forest.Rd
index a7f23e074..7f5e2e604 100644
--- a/man/rand_forest.Rd
+++ b/man/rand_forest.Rd
@@ -5,20 +5,17 @@
 \alias{update.rand_forest}
 \title{General Interface for Random Forest Models}
 \usage{
-rand_forest(mode = "unknown", ..., mtry = NULL, trees = NULL,
-  min_n = NULL, others = list())
+rand_forest(mode = "unknown", mtry = NULL, trees = NULL,
+  min_n = NULL, ...)
 
 \method{update}{rand_forest}(object, mtry = NULL, trees = NULL,
-  min_n = NULL, others = list(), fresh = FALSE, ...)
+  min_n = NULL, fresh = FALSE, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 Possible values for this model are "unknown", "regression", or
 "classification".}
 
-\item{...}{Used for method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{mtry}{An integer for the number of predictors that will
 be randomly sampled at each split when creating the tree models.}
 
@@ -28,18 +25,17 @@ the ensemble.}
 \item{min_n}{An integer for the minimum number of data points
 in a node that are required for the node to be split further.}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{ranger::ranger},
-\code{randomForest::randomForest}, etc.). .}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A random forest model specification.}
 
 \item{fresh}{A logical for whether the arguments should be
 modified in-place of or replaced wholesale.}
 }
-\value{
-An updated model specification.
-}
 \description{
 \code{rand_forest} is a way to generate a \emph{specification} of a model
 before fitting and allows the model to be created using
@@ -54,7 +50,7 @@ that are required for the node to be split further.
 }
 These arguments are converted to their specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to their defaults
+set using the  \code{...} slot. If left to their defaults
 here (\code{NULL}), the values are taken from the underlying model
 functions. If parameters need to be modified, \code{update} can be used
 in lieu of recreating the object from scratch.
@@ -67,14 +63,31 @@ following \emph{engines}:
 \item \pkg{Spark}: \code{"spark"}
 }
 
-Main parameter arguments (and those in \code{others}) can avoid
+Main parameter arguments (and those in \code{...}) can avoid
 evaluation until the underlying function is executed by wrapping the
 argument in \code{\link[rlang:expr]{rlang::expr()}} (e.g. \code{mtry = expr(floor(sqrt(p)))}).
+}
+\note{
+For models created using the spark engine, there are
+several differences to consider. First, only the formula
+interface to via \code{fit} is available; using \code{fit_xy} will
+generate an error. Second, the predictions will always be in a
+spark table format. The names will be the same as documented but
+without the dots. Third, there is no equivalent to factor
+columns in spark tables so class predictions are returned as
+character columns. Fourth, to retain the model object for a new
+R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
+object should be serialized via \code{ml_save(object$fit)} and
+separately saved to disk. In a new session, the object can be
+reloaded and reattached to the \code{parsnip} object.
+}
+\section{Engine Details}{
+
 
 Engines may have pre-set default arguments when executing the
-model fit call. These can be changed by using the \code{others}
+model fit call. These can be changed by using the \code{...}
 argument to pass in the preferred values. For this type of
-model, the template of the fit calls are:
+model, the template of the fit calls are::
 
 \pkg{ranger} classification
 
@@ -105,20 +118,7 @@ constructed using the form \code{estimate +/- z * std_error}. For
 classification probabilities, these values can fall outside of
 \code{[0, 1]} and will be coerced to be in this range.
 }
-\note{
-For models created using the spark engine, there are
-several differences to consider. First, only the formula
-interface to via \code{fit} is available; using \code{fit_xy} will
-generate an error. Second, the predictions will always be in a
-spark table format. The names will be the same as documented but
-without the dots. Third, there is no equivalent to factor
-columns in spark tables so class predictions are returned as
-character columns. Fourth, to retain the model object for a new
-R session (via \code{save}), the \code{model$fit} element of the \code{parsnip}
-object should be serialized via \code{ml_save(object$fit)} and
-separately saved to disk. In a new session, the object can be
-reloaded and reattached to the \code{parsnip} object.
-}
+
 \examples{
 rand_forest(mode = "classification", trees = 2000)
 # Parameters can be represented by a placeholder:
diff --git a/man/surv_reg.Rd b/man/surv_reg.Rd
index a9ee647a6..cca86a6e2 100644
--- a/man/surv_reg.Rd
+++ b/man/surv_reg.Rd
@@ -5,34 +5,28 @@
 \alias{update.surv_reg}
 \title{General Interface for Parametric Survival Models}
 \usage{
-surv_reg(mode = "regression", ..., dist = NULL, others = list())
+surv_reg(mode = "regression", dist = NULL, ...)
 
-\method{update}{surv_reg}(object, dist = NULL, others = list(),
-  fresh = FALSE, ...)
+\method{update}{surv_reg}(object, dist = NULL, fresh = FALSE, ...)
 }
 \arguments{
 \item{mode}{A single character string for the type of model.
 The only possible value for this model is "regression".}
 
-\item{...}{Used for S3 method consistency. Any arguments passed to
-the ellipses will result in an error. Use \code{others} instead.}
-
 \item{dist}{A character string for the outcome distribution. "weibull" is
 the default.}
 
-\item{others}{A named list of arguments to be used by the
-underlying models (e.g., \code{flexsurv::flexsurvreg}). These are not evaluated
-until the model is fit and will be substituted into the model
-fit expression.}
+\item{...}{Other arguments to pass to the specific engine's
+model fit function (see the Engine Details section below). This
+should not include arguments defined by the main parameters to
+this function. For the \code{update} function, the ellipses can
+contain the primary arguments or any others.}
 
 \item{object}{A survival regression model specification.}
 
 \item{fresh}{A logical for whether the arguments should be
 modified in-place of or replaced wholesale.}
 }
-\value{
-An updated model specification.
-}
 \description{
 \code{surv_reg} is a way to generate a \emph{specification} of a model
 before fitting and allows the model to be created using
@@ -43,7 +37,7 @@ model is:
 }
 This argument is converted to its specific names at the
 time that the model is fit. Other options and argument can be
-set using the \code{others} argument. If left to its default
+set using the  \code{...} slot. If left to its default
 here (\code{NULL}), the value is taken from the underlying model
 functions.
 
diff --git a/tests/testthat/test_args_and_modes.R b/tests/testthat/test_args_and_modes.R
index 317663501..a7587a6c2 100644
--- a/tests/testthat/test_args_and_modes.R
+++ b/tests/testthat/test_args_and_modes.R
@@ -1,23 +1,56 @@
 library(testthat)
 library(parsnip)
 library(dplyr)
+library(rlang)
 
 context("changing arguments and engine")
 
 test_that('pipe arguments', {
   mod_1 <- rand_forest() %>%
     set_args(mtry = 1, something = "blah")
-  expect_equal(mod_1$args$mtry, 1)
-  expect_equal(mod_1$others$something, "blah")
-  
-  mod_2 <- rand_forest(mtry = 2, others = list(var = "x")) %>%
+  expect_equal(
+    quo_get_expr(mod_1$args$mtry),
+    1
+  )
+  expect_equal(
+    quo_get_env(mod_1$args$mtry),
+    empty_env()
+  )
+  expect_equal(
+    quo_get_expr(mod_1$others$something),
+    "blah"
+    )
+  expect_equal(
+    quo_get_env(mod_1$others$something),
+    empty_env()
+  )
+
+  x <- 1:10
+  mod_2 <- rand_forest(mtry = 2, var = x) %>%
     set_args(mtry = 1, something = "blah")
-  expect_equal(mod_2$args$mtry, 1)
-  expect_equal(mod_2$others$something, "blah")
-  expect_equal(mod_2$others$var, "x") 
-  
+  expect_equal(
+    quo_get_expr(mod_2$args$mtry),
+    1
+  )
+  expect_equal(
+    quo_get_env(mod_2$args$mtry),
+    empty_env()
+  )
+  expect_equal(
+    quo_get_expr(mod_2$others$something),
+    "blah"
+  )
+  expect_equal(
+    quo_get_env(mod_2$others$something),
+    empty_env()
+  )
+  expect_equal(
+    quo_get_env(mod_2$others$var),
+    global_env()
+  )
+
   expect_error(rand_forest() %>% set_args())
-  
+
 })
 
 
@@ -25,8 +58,8 @@ test_that('pipe engine', {
   mod_1 <- rand_forest() %>%
     set_mode("regression")
   expect_equal(mod_1$mode, "regression")
-  
+
   expect_error(rand_forest() %>% set_mode())
   expect_error(rand_forest() %>% set_mode(2))
   expect_error(rand_forest() %>% set_mode("haberdashery"))
-})
\ No newline at end of file
+})